Practical Guide to DevOps for AI

Practical Guide to DevOps for AI
Date: Fri 12 Jul 2024 - Fri 12 Jul 2024
Register by: 19 June 2024

MLOps focuses on the intersection of ML engineer and Data engineer in combination with existing DevOps practices to streamline model delivery across the machine learning development life cycle. The Objective of MLOps team is to automate the steps of machine learning life cycle into core software system. 

Learning Objectives:

  • What is DevOps? 
  • Popular DevOps tools for Software Development 
  • DevOps for Data Scientists 
  • What is MLOps?
  • Traditional Machine Learning Lifecycle 
  • Challenges in traditional ML lifecycle. 
  • Why MLOps and how MLOps address the challenges? 
  • MLOps Fundamentals 
  • Why DevOps alone is not Suitable for Machine Learning? 
  • Popular MLOps tools for Machine Learning Development
 
MLflow
  • Introduction to MLflow 
  • Components of Mlflow 
  • Mlflow setup 
  • Mlflow tracking component 
  • Mlflow logging functions 
  • Autologging in Mlflow 
  • Tracking server of Mlflow 
  • Mlflow model component 
  • Mlflow model evaluation 
  • Mlflow model registry component 
  • Mlflow Project component 
  • End-to-End Project.
Kubeflow
  • Introduction to Kubernetes 
  • Understanding Containers 
  • Understanding Nodes and Control Plane 
  • Kubernetes API 
  • Kubernetes architecture 
  • Kubectl 
  • Kubernetes Deployments 
  • Kubernetes Pod Networking 
  • Kubernetes Storage and Volumes 
  • Introduction to Kubeflow 
  • Kubeflow components 
  • Kubeflow Pipelines 
  • Training ML Model with Kubeflow 
  • Kubeflow Pipeline 
  • Hyperparameter tuning with Katib 
  • KFServing 
  • End-to-End Project

Pre-requisites

Create a free account to our Training Portal to register for a course and browse all available training courses.

Register now

Join Newsletter

Provide your details to receive regular updates from the STFC Hartree Centre.